This internship is concerned in modeling and testing an information retrieval (IR) system. These systems search in the content of a large collection of documents and retrieve the ones that are relevant to a user's demand, which is usually represented in the form of a query. The content of a document is generally indexed by terms, users' queries are also indexed by terms that are used to identify topics of interest. Terms in documents and queries are usually given specic weights in accordance to their signicance. An information retrieval system has a matching function that assigns scores to documents, each score represents a degree of relevance between a document and a query. Documents are then ranked according to their scores, and top n-documents are retrieved to the user. The subject of the internship aims to exploit fuzzy logic concepts as well as language modeling approaches for IR in order to build and test an information retrieval model.